Mathematical Reasoning on MATH500 (Acc, TF, Acc/TF)
83.8AccASC
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| ASCModel=GEMMA-3-27B2026.01 | 83.8 | 121.29 | 0.69 | |
| ReASC (offline)Model=GEMMA-3-27B2026.01 | 83.6 | 101.32 | 0.83 | |
| ReASC (online)Model=GEMMA-3-27B2026.01 | 83.6 | 100.62 | 0.83 | |
| SCModel=GEMMA-3-27B, k=162026.01 | 82.6 | 291.36 | 0.28 | |
| ESCModel=GEMMA-3-27B, w=42026.01 | 82.6 | 126.56 | 0.65 | |
| ReASC (offline)Model=QWEN-2.5-7B2026.01 | 81.2 | 29.26 | 2.78 | |
| ReASC (online)Model=QWEN-2.5-7B2026.01 | 81.2 | 30.74 | 2.64 | |
| ASCModel=QWEN-2.5-7B2026.01 | 80.8 | 37.25 | 2.17 | |
| SCModel=QWEN-2.5-7B, k=162026.01 | 80.6 | 71.59 | 1.13 | |
| ESCModel=QWEN-2.5-7B, w=42026.01 | 80.6 | 39.71 | 2.03 | |
| pass@1Model=GEMMA-3-27B2026.01 | 77.6 | — | — | |
| SCModel=QWEN-2.5-3B, k=162026.01 | 72.6 | 31.76 | 2.29 | |
| ESCModel=QWEN-2.5-3B, w=42026.01 | 72.6 | 21.75 | 3.34 | |
| ASCModel=QWEN-2.5-3B2026.01 | 72.6 | 20.84 | 3.48 | |
| ReASC (online)Model=QWEN-2.5-3B2026.01 | 72.2 | 17.72 | 4.07 | |
| ReASC (offline)Model=QWEN-2.5-3B2026.01 | 71.8 | 16.8 | 4.27 | |
| SCModel=GEMMA-3-4B, k=162026.01 | 71.6 | 50.15 | 1.43 | |
| ESCModel=GEMMA-3-4B, w=42026.01 | 71.6 | 30.48 | 2.35 | |
| ASCModel=GEMMA-3-4B2026.01 | 71.6 | 28.68 | 2.5 | |
| ReASC (offline)Model=GEMMA-3-4B2026.01 | 71.4 | 25.59 | 2.79 | |
| ReASC (online)Model=GEMMA-3-4B2026.01 | 71.4 | 26.17 | 2.73 | |
| pass@1Model=QWEN-2.5-7B2026.01 | 71.2 | — | — | |
| pass@1Model=GEMMA-3-4B2026.01 | 63.2 | — | — | |
| pass@1Model=QWEN-2.5-3B2026.01 | 61.8 | — | — | |
| ASCModel=LLAMA-3.2-3B2026.01 | 55 | 20.13 | 2.73 | |
| ReASC (offline)Model=LLAMA-3.2-3B2026.01 | 55 | 18.27 | 3.01 | |
| SCModel=LLAMA-3.2-3B, k=162026.01 | 54.8 | 25.31 | 2.17 | |
| ESCModel=LLAMA-3.2-3B, w=42026.01 | 54.8 | 21.41 | 2.56 | |
| ReASC (online)Model=LLAMA-3.2-3B2026.01 | 53.4 | 16.75 | 3.19 | |
| pass@1Model=LLAMA-3.2-3B2026.01 | 41.4 | — | — |